Build Luxury E-Commerce Counterfeit Triage via Mastra AI and Oracle OCI
High-value fashion platforms like Pernia's Pop-Up Shop require strict verification to prevent counterfeit listings and return fraud. Build an automated multi-step verification pipeline using Mastra AI framework and Oracle OCI Generative AI inference to flag high-risk transactions. Achieve a 94% precision score on counterfeit item detection.
What you are building
The core problem, expected build, and operating context for this challenge.
Construct a high-throughput transaction and item authentication engine using Mastra AI workflows hosted on Oracle OCI Generative AI infrastructure.
How work is evaluated
Evaluates item authenticity classification precision against expert luxury appraiser tags.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
fraud_action_check
Ensures high-risk items trigger immediate hold status
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
detection_precision
Precision of flagged counterfeit listings • target: 0.94 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Build stateful workflows using TypeScript with Mastra AI agent utilities
Integrate Oracle OCI Generative AI service endpoints for scalable model inference
Process high-resolution luxury product metadata and customer purchase patterns
Enforce step-by-step risk evaluation schemas for automated merchant hold decisions
Reference links and supporting material
Luxury fashion listing metadata, pricing records, and merchant profile vectors.
Challenge setup
Evaluates item authenticity classification precision against expert luxury appraiser tags.
Challenge input
JSON object with luxury item metadata and seller history
Mastra AI
Provides stateful agent step execution and TypeScript graph orchestrations.
Oracle OCI Generative AI
Provides enterprise cloud runtime for LLM inference calls.
Evaluated output
JSON object with risk score and action flag
- Ensures high-risk items trigger immediate hold status
- Precision of flagged counterfeit listings • target: 0.94 • range: 0-1
- Benchmark: E-Commerce Fraud Detection Benchmark
- Detection Precision target: 0.94
- 1 public reference case
- JavaScript execution harness
- javascript sandbox (unavailable on Versalist)
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
- Mastra AI · Required
- Dora · Optional
- Arcee AI · Optional
- Mastra AI · Required
Evaluation contract
- fraud_action_check · Weight 1
- detection_precision · Weight 1
Metrics
- detection_precision · Target 0.94
Benchmarks
- E-Commerce Fraud Detection Benchmark
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON object with luxury item metadata and seller history
Output: JSON object with risk score and action flag
Execution harness: JavaScript
Evaluation tools
- Mastra AI · Workflow Engine
- Oracle OCI Generative AI · Inference Infrastructure
Recipe freeze marker
No recipe freeze time is recorded.